** Online Appendix for "Vote-Buying, Anti-Corruption Campaigns, and Identity in African Elections"**
** Article authors: Kristen Kao, Ellen Lust, and Lise Rakner **
** This file was written by Kristen Kao, corresponding author: 
** Contact: kristenkao@gmail.com **

use "/Users/xkaokr/Library/Mobile Documents/com~apple~CloudDocs/Projects/Vote Buying Conjoint Malawi 2016/data/Malawi 2016 Vote Buying Dataset.dta"


*Basic model, Figure 1, rating scaled to range between 0-1

egen rating_std = std01(rating)
ta rating_std

eststo clear
eststo OLS: reg rating_std i.appeal2 i.local i.coethnic, cluster(TA)

coefplot OLS, drop( _cons) xline(0) msymbol(circle) mcolor(black) mfcolor(white) levels(95) mlabel mlabposition(12) format(%9.2f) xtitle({bf:Change: Likelihood of Vote (0-1 Scale})) headings(2.appeal2 = "{bf:Candidate Appeal}" 1.local = "{bf:Candidate Attributes}" ) scheme(plotplain) grid(between glcolor(white))


*Online Appendix

eststo Logit: ologit rating i.appeal2 i.local i.coethnic, cluster(TA)

**Model with controls shows same basic outcomes

eststo OLS_controls: reg rating_std i.appeal2 i.local i.coethnic  i.gender i.age_groups i.educat i.elf_quartile i.Region, cluster(TA)

**Table A1 in Appendix 

esttab OLS Logit OLS_controls using "Models_VB2.tex", mtitles() drop() star(+ 0.10 * 0.05 ** 0.01 *** 0.001) collabels(none) label stats(r2  N, fmt(%9.4f %9.0f %9.0fc) labels("R-squared" )) plain b(%9.4f) se(%9.4f) noabbrev se nonumbers lines parentheses replace fragment

*Distribution of original outcome variable 1-4 scale Table A2

ta rating

*Distribution of ratings by appeal Table A3
ta appeal2 rating, row 

*Balance Tables Table B1

eststo Gender: reg gender i.appeal2 i.local i.coethnic, robust

eststo Education: reg educat i.appeal2 i.local i.coethnic if educat != 96 & educat != 98, robust

eststo Wealth: reg savings i.appeal2 i.local i.coethnic, robust

esttab Gender Education Wealth using "balance_VB.tex", mtitles() drop(_cons ) star(+ 0.10 * 0.05 ** 0.01 *** 0.001) collabels(none) label stats(r2  N, fmt(%9.4f %9.0f %9.0fc) labels("R-squared" )) plain b(%9.4f) se(%9.4f) noabbrev se nonumbers lines parentheses replace fragment


**Self-Reported Needs distribution Table B2
ta savings

ta savings if rating!=.

**Asset index distribution Table B3
ta items_MCA1d_quartile

ta items_MCA1d_quartile if rating !=.

**Gender distribution Table B4
ta gender

ta gender if rating!=.

*Education distribution Table B5
ta educat

ta educat if rating !=.

*ELF distribution Table B6
ta elf_quartile

ta elf_quartile if rating !=.

*Region distribution Table B7

ta Region

ta Region if rating !=.

*Received gifts food or money from a candidate Table B8
ta q97

ta q97 if rating !=.

** Heterogeneous (Interaction) effects Table B9

eststo Interactions: reg rating_std i.appeal2##(i.coethnic i.local) i.coethnic##i.local, cluster(TA)

esttab Interactions using "Interactions_VB.tex", mtitles() drop(_cons ) star(+ 0.10 * 0.05 ** 0.01 *** 0.001) collabels(none) label stats(r2  N, fmt(%9.4f %9.0f %9.0fc) labels("R-squared" )) plain b(%9.4f) se(%9.4f) noabbrev se nonumbers lines parentheses replace fragment

**Other analyses (discussed in paper)

*Candidates pass out goods or promise targeted services during elections
ta q95

*Priorities of MPs
ta q227

*Importance of ethnic group member election
ta q210

*Obligation to help coethnics
ta q634_5

*Obligation to help strangers
ta q634_8

*Importance of upholding reputation of ethnic group
ta q222

*MPs respond more quickly to coethnics
ta q224_1

*MPs respond more quickly to voters
ta q224_3

*Obligation to help co-locals
ta q634_6

*MPs respond more quickly to co-locals
ta q224_2

*Importance of co-local election
ta q211

*Obligation to vote for immediate goods
ta q96

*Obligation to vote for future goods
ta q656

*Immediate goods demonstrate ability of candidate to provide for community
ta q657_3

*Future goods demonstrate ability of candidate to provide for community
ta q658_3

*Personally recieved immediate goods in elections
ta q97

*Elections are more than just about immediate handouts
ta q657_2

*Elections are more than just about future handouts
ta q658_2

*Immediate goods corrupt the democratic process
ta q657_1

*Future goods corrupt the democratic process
ta q658_1









